We know when CME are coming but we lack any reliable ways to predict how geoeffective they will be. That's mostly a function of the magnetic handedness of the a

We know when CME are coming but we lack any reliable ways to predict how geoeffective they will be. That's mostly a function of the magnetic handedness of the active region it comes from and if it will arrive at earth with their magnetic field in the right orientation (downwards out of the ecliptic plane, -Bz) to efficiently reconnect with the Earth's magnetic field. There's been many a time when some large sequence of strong CME come and everyone expects intense geomagnetic activity but there's no -Bz so it just buffets things for a bit and causes auroral oval effects. Whereas other times some small lone barely CME will arrive with 7 hours of -20nT Bz and causes 45N aurora. What I'm saying is that the people in charge of chosing to shut down parts of the grid are not ever going to get a confident assertion that damage will happen if they don't. There will be enough uncertainty that this kind of pre-emptive action days in advance will not happen. At best we'll get a few hours of time to mitigate when the events arrive at L1 and the solar monitoring spacecraft there like ACE give actual warning.